search

yylgoodlucky / HDTR-Net

A Real-Time High-Definition Teeth Restoration Network for ArbitraryTalking Face Generation Methods
129 stars 11 forks source link

What is the difference between this model and other SR models? #16

Open Elsaam2y opened 4 months ago

Elsaam2y commented 4 months ago

Hi, I have few questions for better understanding.

1- What is the difference between this model and other SR models? Is it mainly about using the mouth mask? 2- How to run it in real-time? It is too slow on GPU. 3- How to possibly use the reference frame during inference?

Thanks

yylgoodlucky commented 4 months ago

Thank you for your attention to this project. For your question, it’s all explained very clearly in the paper. 1- The difference from the sr model is mainly reflected in inter-frame continuity, real-time and texture restoration. 2- The single frame processing time can reach real-time when we test our model in V100 GPU, and the experimental data is also given. 3- frame by frame. If you have any other questions, you can first read the paper for this project.